Hi Clytie, The phenomenon looks like an issues 28839, which OOo Japanese web pages once encountered.
I would like to recommend testing the phenomena among several Web browsers such as Internet Explorer, Firefox, Safari, and so on, and comparing its results. Vietnamese seems to have some types of encoding as Japanese does. Internet Explorer seems to capable of automatic encoding recognition that determined an encoding used in a web page out of certain predefined set of encodings. But, Internet Explorer 6's specific language targeting auto-recognition, unfortunately did not work with UTF-8 at the moment, several years ago. That is mentioned in the issue above. Internet Explorer 6 Japanese seemed to apply the same encoding used in a current web page (a) to the next web page (b). Where the web page (a) has a link to (b). Consequently, if encodings of (a) and (b) are differ, the web page (b), in a certain condition, gets rendered with garbage characters. Sometimes, the problem seems to temporarily disappears after clicking on a Reload button. An automatic encoding recognition mechanism might make things a little bit complex. At the moment of the issue, explicitly specifying one of Japanese encoding results better than choosing Japanese auto-recognition. It seemed that when IE6 recognizes the web page is encoded in UTF-8 through a HTTP header or <META> tag, it simply ignores the encoding explicitly specified by a user and decides to use UTF-8. In comparison, The same IE6 worked unexpectedly in this way: if Japanese specific auto recognition is chosen by a user, the IE6 seems to try to find out an encoding from Japanese variants and seems to completely forget about other encoding, including UTF-8. Currently, Firefox 3.0.1 lists the followings in the menu: View - Character Encoding - More Encodings - SE & SW Asian - - Vietnamese (TCVN) - Vietnamese (VISC II) - Vietnamese (VIP) - Vietnamese (Windows-1258) There are four local encodings. That implies that Internet Explorer for Vietnamese probably is enforced with auto-encoding-recognition mechanism as well. And there is possibility that the mechanism does not simply work with UTF-8 well, and might have been already solved or would be solved by vendor's updates. As of 6:30 am UTC, September 15, 2008, The Vietnamese web page [1]: - Internet Explorer 6 Japanese (Encode: Unicode (UTF-8)) : Terrible. - Firefox 7.0.1 English (Auto-Detect: East Asian) : Fine. [1] http://vi.openoffice.org/about-downloads.html Ciao, Tora --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
